Understanding the Roles of Self-Esteem, Self-Compassion, and Fear of Self-Compassion in Eating Disorder Pathology: An Examination of Female Students and Eating Disorder Patients Allison C. Kelly, Kiruthiha Vimalakanthan, Jacqueline Carter PII: DOI: Reference:
S1471-0153(14)00050-6 doi: 10.1016/j.eatbeh.2014.04.008 EATBEH 795
To appear in:
Eating Behaviors
Received date: Revised date: Accepted date:
3 February 2014 8 April 2014 30 April 2014
Please cite this article as: Kelly, A.C., Vimalakanthan, K. & Carter, J., Understanding the Roles of Self-Esteem, Self-Compassion, and Fear of Self-Compassion in Eating Disorder Pathology: An Examination of Female Students and Eating Disorder Patients, Eating Behaviors (2014), doi: 10.1016/j.eatbeh.2014.04.008
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
PT
Understanding the Roles of Self-Esteem, Self-Compassion, and
RI
Fear of Self-Compassion in Eating Disorder Pathology:
NU
SC
An Examination of Female Students and Eating Disorder Patients
MA
Allison C. Kellya, Kiruthiha Vimalakanthanb, & Jacqueline Carterc
Department of Psychology, University of Waterloo,
TE
D
a
Waterloo, Ontario, Canada, N2L 3G1
AC CE P
[email protected]
b
Department of Psychology, University of Waterloo, Waterloo, Ontario, N2L 3G1
c
[email protected]
Department of Psychology, Memorial University of Newfoundland St. John’s, Newfoundland, A1B 3X9 [email protected]
1
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
2
Abstract The present study examined the relative contributions of self-compassion, fear of
PT
self-compassion, and self-esteem in eating disorder pathology. One-hundred and fifty-
RI
five female undergraduate students and 97 females entering eating disorders treatment completed the Self-Compassion Scale, Fears of Compassion Scale, Rosenberg Self-
SC
Esteem Inventory, and Eating Disorder Examination Questionnaire. T-tests revealed that
NU
the patient group had lower mean self-compassion and higher mean fear of selfcompassion than the student group. When controlling for self-esteem, high fear of self-
MA
compassion emerged as the strongest predictor of eating disorder pathology in the patient group, whereas low self-compassion was the strongest predictor in the student group.
TE
D
These preliminary results suggest that targeting fear of self-compassion may be important when intervening with individuals suffering from an eating disorder, whereas building
AC CE P
self-compassion may be a valuable approach for eating disorders prevention.
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
3
Understanding the Roles of Self-Esteem, Self-Compassion, and Fear of Self-Compassion in Eating Disorder Pathology:
PT
An Examination of Female Students and Eating Disorder Patients
RI
A large body of research has found that high self-esteem protects against eating disorder pathology whereas low self-esteem is a risk factor for later disturbances in eating
SC
and body image (e.g., Cervera et al., 2002; Gilbert & Meyer, 2005; Granillo, Jones-
NU
Rodriguez, & Carvajal, 2005). Self-esteem is a positive global appraisal of one’s selfworth (Rosenberg, 1965). Although it is associated with many benefits, self-esteem has
MA
been linked to various maladaptive qualities including narcissism, illusory positive beliefs about one’s self and future, and defensiveness in the face of failure (Blaine &
TE
D
Crocker, 1993; Fitch, 1970; Neff & Vonk, 2009). As a result, researchers and clinicians have become interested in self-compassion.
AC CE P
Self-compassion is defined as the tendency to respond to one’s suffering by: adopting an attitude of caring and kindness rather than judgment; viewing one’s pain as common within humanity rather than as isolating; and being mindful of one’s inadequacies rather than ruminating on failures (Neff, 2003a). Perhaps because it is not evaluation-based, self-compassion appears to be a more stable and unconditional form of self-regard than self-esteem (Neff & Vonk, 2009). Although it is moderately positively correlated with self-esteem, it is unrelated to narcissism. It is also associated with acknowledging one’s role in setbacks, as well as learning from and improving upon one’s mistakes (Breines & Chen, 2012; Leary, Tate, Adams, Allen, & Hancock, 2007). Selfcompassion contributes uniquely to various indicators of well-being controlling for selfesteem (Neff, 2003a).
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
4
Self-compassion also appears to promote less maladaptive body- and eatingrelated behaviour. In student samples, it has been associated with more intuitive eating,
PT
fewer body image concerns, and less eating-related guilt controlling for self-esteem
RI
(Schoenefeld & Webb, 2013; Wasylkiw, MacKinnon, & MacLellan, 2012). Both trait and state-induced self-compassion have also been linked to less binge eating (Adams &
SC
Leary, 2007; Webb & Forman, 2012). Ferreira, Pinto-Gouveia, and Duarte (2013) studied
NU
eating disorder patients and community adults and found that self-compassion was associated with a lower drive for thinness in both groups. Finally, Kelly, Carter, and
MA
Borairi (2014) found that eating disorder patients who became more self-compassionate early in treatment had a better treatment response over 12 weeks. These studies reveal
TE
D
that self-compassion may protect against, and facilitate remission of, eating disorder symptoms.
AC CE P
Although self-compassion appears to offer protective effects in the realm of eating and body image, receiving compassion—from others or oneself—appears to be a frightening experience for certain individuals (Gilbert, McEwan, Matos, & Rivis, 2011). People who have a higher fear of self-compassion, due to feeling undeserving of compassion and worrying about lowered personal standards, struggle with lower selfcompassion and more severe psychopathology (Gilbert et al., 2011). Kelly et al. (2013) found that among eating disorder patients who had lower self-compassion, treatment response was poor only if they also had a high fear of self-compassion. This preliminary study suggests it is not simply the shortage of self-compassion, but also the presence of fear of self-compassion that may be particularly deleterious with respect to eating disorder pathology.
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
5
Despite the associations between self-compassion, eating behaviour, and body image, there have yet to be systematic tests of the relative contributions of self-esteem,
PT
self-compassion, and fear of self-compassion to eating disorder pathology using
RI
comprehensive measures of eating disorder symptomatology; and examining both clinical and non-clinical samples. The present study sought to fill this gap in the literature by
SC
testing the following hypotheses. First, female eating disorder patients would have lower
NU
mean self-compassion and higher mean fear of self-compassion than female students. Second, controlling for body mass index (BMI) and self-esteem, lower trait self-
MA
compassion and higher fear of self-compassion would uniquely predict eating disorder
Participants
Method
TE
D
pathology in both groups.
AC CE P
The student sample consisted of 155 female undergraduates (mean age=20; SD=5.0) recruited from a university participant pool. Ethnicity was: 48.3% Caucasian, 19.4% South Asian, 12.9% East Asian, 6.5% Southeast Asian, 3.2% Black/African, 3.2% bi-racial, 2.6% West Indian/Caribbean, and 5% other. Mean BMI was 23.08 (SD=4.99). The patient sample consisted of 97 females (mean age=28; SD=9.6) beginning treatment at Toronto General Hospital’s eating disorders program. All met DSM-IV-TR criteria for an eating disorder (29.6% bulimia nervosa, 27.2% anorexia nervosa restricting type, 18.5% anorexia nervosa binge-purge type, 24.7% eating disorder not otherwise specified). Ethnicity was: 79.2% Caucasian, 10.8% Hispanic, 4.5% East Asian, 2.8% African-Canadian, and 2.9% other. Mean BMI was 20.99 (SD=5.57).
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
6
Measures Eating Disorder Examination Questionnaire (EDE-Q; Fairburn, 2008). The EDE-
PT
Q is a 36-item questionnaire measure of eating disorder symptomatology, with four
RI
different subscales: Shape Concern, Weight Concern, Eating Concern, and Dietary Restraint. The mean of these subscales yields a composite global score. The EDE-Q has
SC
good test-retest reliability (Luce & Crowther, 1999). The Cronbach’s alpha in both our
NU
samples was .95, indicating strong internal consistency.
Self-Compassion Scale – Short Form (SCS-SF; Raes, Pommier, Neff, & Van
MA
Gucht, 2011). The SCS-SF is a 12-item adaptation of the 26-item SCS (Neff, 2003b), with which it correlates near perfectly. The SCS-SF assesses how participants typically
D
respond to themselves at times of distress or failure. Sample items include: “I’m tolerant
TE
of my own flaws and inadequacies”, and “I’m kind to myself when I’m experiencing
AC CE P
suffering”. The SCS-SF has good internal consistency; Cronbach’s alphas were 0.92 and 0.85 in the student and patient samples, respectively. Fears of Compassion Scale (FCS; Gilbert, McEwan, Matos, & Rivis, 2011). Fear of self-compassion was assessed with the 15-item section of the FCS assessing fears of expressing kindness and compassion toward oneself. Sample items include: “I feel that I don’t deserve to be kind and forgiving to myself” and “I fear that if I become kinder and less self-critical to myself then my standards will drop.” The FCS demonstrates excellent internal consistency, with a Cronbach’s alpha of 0.95 in both samples. Rosenberg Self-Esteem Scale (RSE; Rosenberg, 1965). The RSE is a widely-used 10-item measure of self-esteem (e.g., “I take a positive attitude towards myself”). It has high test-retest reliability (Robins, Hendin, & Trzesniewski, 2001) and good internal
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
7
consistency. Individual item scores were unavailable for our patient sample, but the Cronbach’s alpha was 0.81 in our student sample.
PT
Results
RI
Patterns of Missing Data
In the patient sample, BMI, eating disorder symptomatology, and self-esteem
SC
were assessed as part of a separate assessment package; complete data were obtained for
for all other patient and student measures.
MA
Analytic Strategy
NU
75, 82, and 76 out of the 97 participants, respectively. Near-complete data were available
First, t-tests examined group differences between our student and patient samples
TE
D
on study variables. Equality of variances tests revealed that variances between the two groups only differed for fear of self-compassion, F (93, 150) = 1.74, p < .01. We
AC CE P
therefore present Satterthwaite t-test results when comparing group means on this variable and pooled t-tests for all other variables. Second, multiple regressions investigated the unique contributions of self-compassion and fear of self-compassion to EDE-Q global and subscale scores, while controlling for self-esteem and BMI. All predictor variables were standardized to facilitate interpretation of the results. Between-Group Differences in Study Variables Means and standard deviations for all variables within each group are presented in Table 1. T-tests revealed between-group differences. The patient sample had a lower mean BMI, p
S1471-0153(14)00050-6 doi: 10.1016/j.eatbeh.2014.04.008 EATBEH 795
To appear in:
Eating Behaviors
Received date: Revised date: Accepted date:
3 February 2014 8 April 2014 30 April 2014
Please cite this article as: Kelly, A.C., Vimalakanthan, K. & Carter, J., Understanding the Roles of Self-Esteem, Self-Compassion, and Fear of Self-Compassion in Eating Disorder Pathology: An Examination of Female Students and Eating Disorder Patients, Eating Behaviors (2014), doi: 10.1016/j.eatbeh.2014.04.008
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
PT
Understanding the Roles of Self-Esteem, Self-Compassion, and
RI
Fear of Self-Compassion in Eating Disorder Pathology:
NU
SC
An Examination of Female Students and Eating Disorder Patients
MA
Allison C. Kellya, Kiruthiha Vimalakanthanb, & Jacqueline Carterc
Department of Psychology, University of Waterloo,
TE
D
a
Waterloo, Ontario, Canada, N2L 3G1
AC CE P
[email protected]
b
Department of Psychology, University of Waterloo, Waterloo, Ontario, N2L 3G1
c
[email protected]
Department of Psychology, Memorial University of Newfoundland St. John’s, Newfoundland, A1B 3X9 [email protected]
1
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
2
Abstract The present study examined the relative contributions of self-compassion, fear of
PT
self-compassion, and self-esteem in eating disorder pathology. One-hundred and fifty-
RI
five female undergraduate students and 97 females entering eating disorders treatment completed the Self-Compassion Scale, Fears of Compassion Scale, Rosenberg Self-
SC
Esteem Inventory, and Eating Disorder Examination Questionnaire. T-tests revealed that
NU
the patient group had lower mean self-compassion and higher mean fear of selfcompassion than the student group. When controlling for self-esteem, high fear of self-
MA
compassion emerged as the strongest predictor of eating disorder pathology in the patient group, whereas low self-compassion was the strongest predictor in the student group.
TE
D
These preliminary results suggest that targeting fear of self-compassion may be important when intervening with individuals suffering from an eating disorder, whereas building
AC CE P
self-compassion may be a valuable approach for eating disorders prevention.
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
3
Understanding the Roles of Self-Esteem, Self-Compassion, and Fear of Self-Compassion in Eating Disorder Pathology:
PT
An Examination of Female Students and Eating Disorder Patients
RI
A large body of research has found that high self-esteem protects against eating disorder pathology whereas low self-esteem is a risk factor for later disturbances in eating
SC
and body image (e.g., Cervera et al., 2002; Gilbert & Meyer, 2005; Granillo, Jones-
NU
Rodriguez, & Carvajal, 2005). Self-esteem is a positive global appraisal of one’s selfworth (Rosenberg, 1965). Although it is associated with many benefits, self-esteem has
MA
been linked to various maladaptive qualities including narcissism, illusory positive beliefs about one’s self and future, and defensiveness in the face of failure (Blaine &
TE
D
Crocker, 1993; Fitch, 1970; Neff & Vonk, 2009). As a result, researchers and clinicians have become interested in self-compassion.
AC CE P
Self-compassion is defined as the tendency to respond to one’s suffering by: adopting an attitude of caring and kindness rather than judgment; viewing one’s pain as common within humanity rather than as isolating; and being mindful of one’s inadequacies rather than ruminating on failures (Neff, 2003a). Perhaps because it is not evaluation-based, self-compassion appears to be a more stable and unconditional form of self-regard than self-esteem (Neff & Vonk, 2009). Although it is moderately positively correlated with self-esteem, it is unrelated to narcissism. It is also associated with acknowledging one’s role in setbacks, as well as learning from and improving upon one’s mistakes (Breines & Chen, 2012; Leary, Tate, Adams, Allen, & Hancock, 2007). Selfcompassion contributes uniquely to various indicators of well-being controlling for selfesteem (Neff, 2003a).
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
4
Self-compassion also appears to promote less maladaptive body- and eatingrelated behaviour. In student samples, it has been associated with more intuitive eating,
PT
fewer body image concerns, and less eating-related guilt controlling for self-esteem
RI
(Schoenefeld & Webb, 2013; Wasylkiw, MacKinnon, & MacLellan, 2012). Both trait and state-induced self-compassion have also been linked to less binge eating (Adams &
SC
Leary, 2007; Webb & Forman, 2012). Ferreira, Pinto-Gouveia, and Duarte (2013) studied
NU
eating disorder patients and community adults and found that self-compassion was associated with a lower drive for thinness in both groups. Finally, Kelly, Carter, and
MA
Borairi (2014) found that eating disorder patients who became more self-compassionate early in treatment had a better treatment response over 12 weeks. These studies reveal
TE
D
that self-compassion may protect against, and facilitate remission of, eating disorder symptoms.
AC CE P
Although self-compassion appears to offer protective effects in the realm of eating and body image, receiving compassion—from others or oneself—appears to be a frightening experience for certain individuals (Gilbert, McEwan, Matos, & Rivis, 2011). People who have a higher fear of self-compassion, due to feeling undeserving of compassion and worrying about lowered personal standards, struggle with lower selfcompassion and more severe psychopathology (Gilbert et al., 2011). Kelly et al. (2013) found that among eating disorder patients who had lower self-compassion, treatment response was poor only if they also had a high fear of self-compassion. This preliminary study suggests it is not simply the shortage of self-compassion, but also the presence of fear of self-compassion that may be particularly deleterious with respect to eating disorder pathology.
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
5
Despite the associations between self-compassion, eating behaviour, and body image, there have yet to be systematic tests of the relative contributions of self-esteem,
PT
self-compassion, and fear of self-compassion to eating disorder pathology using
RI
comprehensive measures of eating disorder symptomatology; and examining both clinical and non-clinical samples. The present study sought to fill this gap in the literature by
SC
testing the following hypotheses. First, female eating disorder patients would have lower
NU
mean self-compassion and higher mean fear of self-compassion than female students. Second, controlling for body mass index (BMI) and self-esteem, lower trait self-
MA
compassion and higher fear of self-compassion would uniquely predict eating disorder
Participants
Method
TE
D
pathology in both groups.
AC CE P
The student sample consisted of 155 female undergraduates (mean age=20; SD=5.0) recruited from a university participant pool. Ethnicity was: 48.3% Caucasian, 19.4% South Asian, 12.9% East Asian, 6.5% Southeast Asian, 3.2% Black/African, 3.2% bi-racial, 2.6% West Indian/Caribbean, and 5% other. Mean BMI was 23.08 (SD=4.99). The patient sample consisted of 97 females (mean age=28; SD=9.6) beginning treatment at Toronto General Hospital’s eating disorders program. All met DSM-IV-TR criteria for an eating disorder (29.6% bulimia nervosa, 27.2% anorexia nervosa restricting type, 18.5% anorexia nervosa binge-purge type, 24.7% eating disorder not otherwise specified). Ethnicity was: 79.2% Caucasian, 10.8% Hispanic, 4.5% East Asian, 2.8% African-Canadian, and 2.9% other. Mean BMI was 20.99 (SD=5.57).
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
6
Measures Eating Disorder Examination Questionnaire (EDE-Q; Fairburn, 2008). The EDE-
PT
Q is a 36-item questionnaire measure of eating disorder symptomatology, with four
RI
different subscales: Shape Concern, Weight Concern, Eating Concern, and Dietary Restraint. The mean of these subscales yields a composite global score. The EDE-Q has
SC
good test-retest reliability (Luce & Crowther, 1999). The Cronbach’s alpha in both our
NU
samples was .95, indicating strong internal consistency.
Self-Compassion Scale – Short Form (SCS-SF; Raes, Pommier, Neff, & Van
MA
Gucht, 2011). The SCS-SF is a 12-item adaptation of the 26-item SCS (Neff, 2003b), with which it correlates near perfectly. The SCS-SF assesses how participants typically
D
respond to themselves at times of distress or failure. Sample items include: “I’m tolerant
TE
of my own flaws and inadequacies”, and “I’m kind to myself when I’m experiencing
AC CE P
suffering”. The SCS-SF has good internal consistency; Cronbach’s alphas were 0.92 and 0.85 in the student and patient samples, respectively. Fears of Compassion Scale (FCS; Gilbert, McEwan, Matos, & Rivis, 2011). Fear of self-compassion was assessed with the 15-item section of the FCS assessing fears of expressing kindness and compassion toward oneself. Sample items include: “I feel that I don’t deserve to be kind and forgiving to myself” and “I fear that if I become kinder and less self-critical to myself then my standards will drop.” The FCS demonstrates excellent internal consistency, with a Cronbach’s alpha of 0.95 in both samples. Rosenberg Self-Esteem Scale (RSE; Rosenberg, 1965). The RSE is a widely-used 10-item measure of self-esteem (e.g., “I take a positive attitude towards myself”). It has high test-retest reliability (Robins, Hendin, & Trzesniewski, 2001) and good internal
ACCEPTED MANUSCRIPT Self-Compassion and Eating Disorder Pathology
7
consistency. Individual item scores were unavailable for our patient sample, but the Cronbach’s alpha was 0.81 in our student sample.
PT
Results
RI
Patterns of Missing Data
In the patient sample, BMI, eating disorder symptomatology, and self-esteem
SC
were assessed as part of a separate assessment package; complete data were obtained for
for all other patient and student measures.
MA
Analytic Strategy
NU
75, 82, and 76 out of the 97 participants, respectively. Near-complete data were available
First, t-tests examined group differences between our student and patient samples
TE
D
on study variables. Equality of variances tests revealed that variances between the two groups only differed for fear of self-compassion, F (93, 150) = 1.74, p < .01. We
AC CE P
therefore present Satterthwaite t-test results when comparing group means on this variable and pooled t-tests for all other variables. Second, multiple regressions investigated the unique contributions of self-compassion and fear of self-compassion to EDE-Q global and subscale scores, while controlling for self-esteem and BMI. All predictor variables were standardized to facilitate interpretation of the results. Between-Group Differences in Study Variables Means and standard deviations for all variables within each group are presented in Table 1. T-tests revealed between-group differences. The patient sample had a lower mean BMI, p
